My Prehistoric Saga: What did Spinosaurus Look Like Part 1: The Limbs.

Quite honestly, there has never been a more irritating (no Irritator), and confusing, dinosaur than Spinosaurus. At least, to me. Every year or two, it seems that paleontologists want to start a war to see which of them knew what Spinosaurus truly looked like, and what its behavior was. To bring this point home, a new paper by Sereno et al., (2022) says that Spinosaurus wasn't aquatic, was bipedal, and brought back the classic semi-circle-shaped sail on its back. So many Spinosaurus publications have been published since 2014, and all I feel at this point is annoyance. I had a much more enjoyable experience studying Dryptosaurus (including the "Nanotyrannus" specimens) than I do studying Spinosaurus. I'm going to end this excruciating experience once and for all and come up with a conclusive answer as to what this animal was doing in its ecology (at least, to the best of my abilities)!

In this series of posts, I'm going to exhaustively study Spinosaurus to see what it might've looked like, and what its behavior was in its ecological niche. Was it bipedal or quadrupedal? Could it swim? What did its sail look like? All of those questions will be answered here! I will say that this is going to be for MY benefit, and I know that I will go against what the majority of the paleontological community thinks. However, I have one motto now: Do your best and follow the evidence. I did this for Dryptosaurus, so I'm going to do this for Spinosaurus.

Part 1: Limb Proportions:
1. A Brief History:
Let's get started by talking about the limbs, and if Spinosaurus was bipedal or quadrupedal. Before we begin, here's a brief history lesson of Spinosaurus' posture:

Originally, Spinosaurus was seen as being either bipedal or quadrupedal. The first reconstruction by Stromer showed Spinosaurus as a typical bipedal carnivorous theropod (Stromer, 1936, p. 65). In the 1970's, Spinosaurus was shown as being quadrupedal (Glut, 2001, p. 82). Heck, it looked like Dimetrodon. Some scientists thought that it was quadrupedal with a bison-humped sail (Hecht, 1996, para. 1) (Black, 2011). Later on, it was bipedal again and this view stayed when Spinosaurus' relative Suchomimus was discovered (Sereno et al., 1998, p. 1300) (Glut, 2001, p. 84). In 2014, Ibrahim et al. discovered a new Spinosaurus skeleton (the now famous neotype specimen FSAC-KK 11888) and said that the taxon was quadrupedal. Duane Nash said that Spinosaurus was a belly-slider, and used its arms to help push its body forward (Nash, 2014, "Did Bakker Get Spinosaurus Right After All?"). Henderson (2018) said that Spinosaurus was bipedal, and that Ibrahim et al., (2014) was wrong. Ibrahim et al., (2020b) said that Spinosaurus was quadrupedal. Hone and Holtz (2021) argued for bipedalism. In 2022, Ibrahim's original team broke up and now there's fighting between them as to who is right about Spinosaurus' posture. Ibrahim joined Fabbri et al., (2022), which said that Spinosaurus was quadrupedal (Figure 1). Sereno, an original member of Ibrahim's first team, made his own team of researchers and wrote a paper arguing for bipedalism (Sereno et al., 2022).

In between all of that, paleo-artists Luis V. Rey and Dr. Scott Hartman came up with their own designs for Spinosaurus. Rey said that Spinosaurus was quadrupedal (Rey, 2014, "What happens when Spinosaurus
runs ashore?") (Rey, 2015, "Popularizing science... The right way!"), while Hartman said that it was bipedal (Hartman, 2020, "The Road to Spinosaurus IV: Not Your Father's JP3 Spinosaurus"). Recently, another paleo-artist named Hank Sharpe made Spinosaurus trend on Twitter by saying that Spinosaurus' femur wasn't capable of holding its weight. This suggests that Spinosaurus wasn't bipedal (Sharpe, 2023,
"Think Spinosaurus' legs look kinda wimpy for its size?...").

Spinosaurus holotype IPHG 1912 VIII 19 from Stromer (1936) (p. 65):

Spinosaurus from 1970's(?) as described by Glut (2001) (p. 82):

Bison-backed Spinosaurus by R. E. Johnson and from a paper by Bailey (1997) (Black, 2011):

Spinosaurus after Ibrahim et al., (2014) by David Bonadonna:

Spinosaurus by Luis V. Rey (Rey, 2014, "What happens when Spinosaurus runs ashore?"):

"Spino tidal" by Duane Nash (Nash, 2014, "Did Bakker Get Spinosaurus Right After All?"):

Spinosaurus' posture from Henderson (2018) (Figure 1):

Scott Hartman's reconstruction of Spinosaurus (Hartman, 2020, "The Road to Spinosaurus IV: Not Your Father's JP3 Spinosaurus"):

Hone's and Holtz's Spinosaurus neotype design by Genya Masukawa (Hone and Holtz, 2021, Figure 1):

Spinosaurus' posture by Sereno et al., (2022) (Figure 1A-E):

2. Math:

Now that that's out of the way, what do I have to say about this? Well, I think I have an answer (for myself, anyway). In Charig and Miller (1997), they were trying to calculate if Baryonyx was a biped or not. The authors stated that the arms of Baryonyx were strong enough for it to have been quadrupedal. However, the humerus was 39% the length of the femur. Compared to Stegosaurus, which had a humerus that was 43-55% the length of its femur (p. 63). Based on the percentages, I believe that Baryonyx was bipedal. Its femur wasn't completely preserved (p. 51), but based on its phylogenetic position with Baryonyx/Suchomimus tenerensis (Sereno et al., 1998, p. 1300 Figure 4) (Fabbri et al., 2022, p. 854 Figure 1), I believe that its femur was longer than its arms and enabled bipedalism. I'm going to use this technique on Spinosaurus, in particular the neotype FSAC-KK 11888.

I got this idea from Goo (2022), which described a couple of new Spinosaurus bones. He places them under the species name Spinosaurus "dorsojuvencus," but I'm ignoring that for the time being. Any spinosauroid bones placed under the genus Spinosaurus will be placed under the species Spinosaurus aegyptiacus for now. Anyway, Goo said that the specimen Alpha Male 9109 had a radius that was 64.7% the length of the Spinosaurus neotype's femur (Restoration of Spinosaurid limb bone, para. 2). He then remarked that Spinosaurus was probably both bipedal and quadrupedal (Order Aliacollisauria: Diagnosis). It should be noted that Goo stated that his paper is outside mainstream paleontology. He is also a chemist that studies private fossils, but he has no background in paleontology (Foreword, para. 2-3). At this point, I am desperate to find an answer for Spinosaurus' weirdness that I'm going to go with anything that I can find. Of course, I will do my best to see if what Goo says matches any other sources that I have. With that in mind, I'm going to use Goo (2022) and Charig and Miller (1997) to see if Spinosaurus was capable of being a biped only.

Spinosaurus:

1.) FSAC-KK 11888 (neotype):

Skull: 123 cm.

Femur: 62.5 cm.

Body: 10.3 m.
Source: Ibrahim et al., (2020b) (Supplementary Materials: Date File 2, Body dimensions, body mass, body segment masses, and whole body center of mass, p. 1):

Originally, Ibrahim et al., (2014) gave an estimated length for the humerus at 51 cm, and an estimated length of 24 cm for the radius. This was based on Suchomimus as well (Supplementary Materials, p. 30 Table S1):

Femur length (also from Table S1 on p. 30):

Humerus-to-femur ratio for the neotype using Ibrahim et al., (2014):
Humerus; 51 cm.
Radius: 24 cm.
Femur: 61 cm.

Humerus:

61 - 51 = 10.
10/61*100 = 16.4% decrease.
100% - 16.4% = 83.6% the length of the femur.

Radius:
24 - 61 = 37.
37/61*100 = 60.7% decrease.
100% - 60.7% = 39.3% the length of the femur

The neotype would've had a humerus that was 83.6% the length of the femur, with the radius being 39.3%. This would've meant that Spinosaurus was a quadrupedal animal. This percentage is larger than Stegosaurus', which is surprising.

Now, using the new estimated skull length for the neotype from Ibrahim et al., (2020b), I'm going to estimate the neotype's humerus and radius length. Then, I'll check to see if the animal was quadrupedal or bipedal.

Humerus and radius estimate based on Suchomimus:

Lengths from Sereno et al., (1998) and Hendrickx et al., (2016):
Skull: 119 cm.
Humerus: 56 cm.
Radius: 25.5 cm.
Femur: 107.5 cm.

Humerus:
(Using skull lengths) 119 - 123 = 4.
4/119*100 = 3.4% increase.

56 cm + 3.4% = 57.9 cm for FSAC's humerus length.

Radius:

25.5 cm + 3.4% = 26.4 cm for radius length.

Humerus-to-femur ratio:

57.9 cm - 62.5 cm = 4.6.

4.6/62.5*100 = 7.4% decrease.

100% - 7.4% = 92.6% the length of the femur.

Radius-to-femur ratio:

26.4 cm - 62.5 cm = 36.1.

36.1/62.5*100 = 57.76% decrease.

100% - 57.76% = 42.24% the length of the femur.

This provides a greater percentage than from Ibrahim et al., (2014)! The humerus was 92.6% the length of its femur, with the radius being 42.24%. This would indicate that it was a quadrupedal animal.

Update (8/24/24):

Just in case, using the femur length from Ibrahim et al., (2020b) with the humerus length from Ibrahim et al., (2014), we get:

62.5 - 51 = 11.5.
11.5/61*100 = 18.9% decrease.
100% - 18.9% = 81.1% the length of the femur.

This would still indicate Spinosaurus as being a quadrupedal animal.

2.) Alpha Male 9109 (Goo, 2022):
Measurements (Goo, 2022, Figure 26) (Note: Squares are in cm in :

Radius (C): ~41.8 cm at best (Alpha Male 9109).

Femur (D): 62 cm (FSAC-KK 11888).

Finding Woo’s 67.7% estimate (radius of Alpha Male compared to femur of neotype):

62 - 41.8 = 20.2.

20.2/62*100 = 32.6% decrease.

100% - 32.6% = 67.4% the total length of the femur.

Guess work for humerus length compared to neotype:

Humerus:

26.4 - 41.8 = 15.4.

15.4/26.4*100 = 58.3% increase.

57.9 cm + 58.3% = 91.7 cm.

Femur:

62.5% + 58.3% = 98.9 cm.

-Body:

10.3 m + 58.3% = 54 feet (16.3 meters).

Length of humerus compared to the femur:

91.7 cm - 98.9 cm = 7.2.

7.2/98.9*100 = 7.3% decrease.

100% - 7.2% = 92.8% the length of the femur.

Looks like Alpha Male 9109 was also quadrupedal, with a humerus that was 92.8% of the femur.

3. Evidence of Pronation in Theropoda?
It has been stated that theropods couldn't pronate their hands, preventing their palms from touching the ground (Hone, 2009, "Theropods are clappers, not slappers," para. 3). However, there seems to be a trace fossil that shows evidence of a theropod placing its hands on the ground with the palm facing the soil (McCrea et al., 2002, Abstract) (Black, 2009). McCrea et al., said that the claw marks were made "simultaneous and parallel strikes or drags of both the animal's forelimbs while it was walking" (para. 1). Black said that this showed that theropods "could hold their arms with elbows out and palms down" (para. 3). More recently, Caneer et al., (2021) found a trace fossil of a T. rex placing its arms and hands onto the ground. The hands were used, along with the legs, to help the animal get up from a prone position
(Abstract; pp. 29-30; p. 33 Figure 6 C; p. 35 Figure 8). So, it seems that theropods were capable of doing something with their hands akin to pronation, if not to a minimum degree. With Spinosaurus and it's wonky design, this could have been turned up a notch to with its hands being fully pronated (perhaps). At the very least, it probably did something similar to the theropods that made the trace fossils from McCrea et al., (2002) and Caneer et al., (2021).

Quick side note: Luis V. Rey said that paleontologist Alan Gishlike showed him that theropods hands could "bend slightly outwards" (Rey, 2015, "Popularizing science... The right way!", para. 4).

Theropod pronating its hands based on McCrea et al., (2002) by Michael Shrepnick (Black, 2009):

T. rex arm trace fossils (Caneer et al., 2021, p. 33 Figure 6, C):

Drawing of T. rex showing how it used its arms to sit down, or even get up from the ground (p. 35 Figure 8):

4. Possible Spinosaurus Manual Ungual:
Something caught my eye regarding Spinosaurus' (hopefully) manual ungual. Ibrahim et al., (2020a) showed a collection of manual unguals (hand claws). The first one, catalogued as NMC 41820 (Figure 111A-B), is labelled as possibly belonging to a spinosauroid. The claw is "gently recurved" compared to others (Theropoda: Manual ungual morphotype 1). If this is a spinosauroid claw, then it more-than-likely belongs to Spinosaurus. It differs for the extremely curved Baryonyx and Suchomimus manual unguals (Charig and Miller, 1997, p. 47 Figure 35) (Hone, 2012, Suchomimus). I suspect that Spinosaurus wasn't using its hands like its two cousins were, and were probably for helping in supporting its weight.

Hone and Holtz, Jr. (2017) stated that spinosaurids used their hands for digging, or even for intimidation, rather than catching and dissecting prey. They would have swallowed their prey (pp. 1128-1129; Figure 6). Given the morphology of NMC 41820, if it does belong to Spinosaurus, then it didn't use its unguals to dig but to help it to walk. Using its claws for intimidation might have worked too.

Baryonyx manual ungual:

Suchomimus arm (Hone, 2012, Suchomimus):

Spinosauroid (Spinosaurus?) manual ungual (Ibrahim et al., 2020a, Figure 111A-B):

Description of NMC 41820 (Ibrahim et al., 2020a, Theropoda: Manual ungual morphotype 1):

Center of Mass Estimation:
The biggest question that I've been trying to answer is: What is Spinosaurus' center of mass (CoM)? Knowing this would help to answer the bipedal or quadrupedal debate once and for all. Originally, Ibrahim et al., (2014) got 1.04 meters for the CoM (Ibrahim et al., 2020b, Supplementary Materials: Data File 2: Body dimensions, body mass, body segment masses, and whole body center of mass, p. 8). Henderson (2018) got 0.3182 (Figure 7), or 0.48 m (Ibrahim et al., 2020b, Supplementary Materials, p. 31: Body mass, segment masses, and centre of mass (CoM)), which made Spinosaurus a bipedal animal. Ibrahim et al., (2020b) got 0.725-0.825 m, which would've made Spinosaurus quadrupedal again (Supplementary Materials, p. 31: Body mass, segment masses, and centre of mass (CoM)). Sereno et al., (2022) got 13.2-28.5 cm (they gave a femoral length of 40 cm), which made Spinosaurus bipedal again (Materials and Methods, Flesh model density, dimensions and properties, para. 4).

I'm going to find the CoM myself, using the CoM equation:

Center of Mass equation:
Xcm = (m1)(x1) + (m2)(x2)/m1 + m2

Equation is from Georgia State University:

m = mass of object.
x = length of object.

Center of Gravity equation:
CoG = (w1)(x1) + (w2)(x2) + ...to end/w1 + w2 + ...to end.

Equation is from Study.com (Couldn't unlock without making an account, so I apologize for the texture):

w = weight of object.
x = length of object.

However, before I begin, I think it needs to be stated that the CoM, and the center of gravity (CoG), equations are "used synonymously" (Georgia State University, Hyperphysics: Center of Mass, para. 1). Therefore, using either the CoM or CoG equation would work. From what I've seen, the CoM equation is typically used to find the CoM/CoG between two points, while the CoG equation is used to find the CoM/CoG between three or more points. The equations for both, on the other hand, are basically the same. I'm going to use the CoG equation because there's more than two points. If the CoM/CoG is larger than the femoral length, then Spinosaurus was quadrupedal (Ibrahim et al., 2014, p. 1615) (Ibrahim et al., 2020b, Supplementary Materials, p. 31: Body mass, segment masses, and centre of mass (CoM)).

Body and mass numbers came from Ibrahim et al., (2020b) (Supplementary Materials: Data File 2: Body dimensions, body mass, body segment masses, and whole body center of mass) that are in the "without air sacs" section. This is the first section:

-2.8 m = head and neck length.
-3.09 m = torso and hip length.
-5.04 m = tail length.
-453 kg = head and neck weight (I added the skull and neck weights from p. 5 because the torso and hip lengths were added together on p. 1).
-2124 kg = torso and hip weight.
-629 kg = tail weight.

Xcm = (453 kg)(2.8 m) + (2124 kg)(3.09 m) + (629 kg)(5.04 m)/453 + 2124 + 629

Xcm = 1,268.4 + 6563.16 + 3,170.16/3206
Xcm = 11,001.56/3206
Xcm = 3.43 m.

CoG = (453 kg)(2.8 m) + (2124 kg)(3.09 m) + (629 kg)(5.04 m)/453 + 2124 + 629
CoG = 1,268.4 + 6563.16 + 3,170.16/3206
CoG = 11,001.56/3206
CoG = 3.43 m.

Note: A length for Spinosaurus' sail wasn't given by the authors, so I left it out.

Update (8/28/23):
I found the CoM equation used by Ibrahim et al., (2020b), so let's try and use it to find the CoM!

CoM equation from Ibrahim et al., (2020b) (Supplementary Materials, p. 30):

Xc = (Xs(Ms - Mzd))/Mt (Supplementary Materials, p. 30: Body mass, segment masses, and centre of mass (CoM)).

-Xs = Cartesian coordinate (Data File 2, p. 4).
-Ms = Body segment mass (Data File 2, p. 5).
-Mzd = Zero-density volume (kg/m^3 or kg.m-3) (Data File 2, p. 5).
-Mt = Total body mass (Data File 2, p. 5).

Cartesian coordinates (Data File 2, p. 4):

Mass estimates using the Cartesian coordinates, and zero-density volume, part 1 (p. 5):

Part 2 (p. 6):

(?)Total body mass:
Ms - Mzd
= 3865 kg - 1000 kg/m^3 (3865 = Add up all body segments on p. 5; or, is 3.864.7 kg on p. 5).
= 2865 kg (= Mt).

-*Or, Mt body mass is 3864.7 kg (without air sacs) on p. 5.

Cartesian coordinate: 6.2468850742257 (Data File 2, p. 4).

1.): Body mass is 2865 kg:

Xc = (6.2468850742257(3865 kg - 1000 kg/m^3))/2865

Xc = (6.2468850742257 (2865))/2865
Xc = 17897.3257377/2865
Xc = 4.63 m

2.) Cartesian coordinate is 6.2468850742257 and body mass is 3864.7 kg:

Xc = (6.2468850742257(3864.7 kg - 1000 kg/m^3))/3864.7 kg

Xc = (6.2468850742257(2864.7))/3864.7
Xc = 17895.4516721/3864.7
Xc = 4.63 m

3.) Cartesian coordinate is 6.2468850742257, numerator body mass is 3865 kg, and denominator body mass is 3864.7 kg:

Xc = (6.2468850742257(3865 kg - 1000 kg/m^3))/3864.7

Xc = (6.2468850742257(2865))/3864.7
Xc = 17897.3257377/3864.7
Xc = 4.63 m

Torso segments:

*Torso segment part 1 (No air sacs):

Cartesian coordinate is 6.9071243 (Data File 2, p. 4) and body mass is 3864.7 (Data File 2, p. 5):

Xc = (6.9071243(1734 kg - 1000 kg/m^3))/3864.7 kg

Xc = (6.9071243(743))/3864.7
Xc = 5131.99335/3864.7
Xc = 1.32791506 m (or 1.33 m).

**Torso segment part 2 (No air sacs):

Cartesian coordinate is 6.2468850742257 and body mass is 3864.7:

Xc = (6.2468850742257(1734 kg - 1000 kg/m^3))/3864.7

Xc = (6.2468850742257(743))/3864.7
Xc = 4641.43561015/3864.7
Xc = 1.201 m.

-Torso numbers give better results!

Torso segment part 3 (Segment total mass - air space mass):

Cartesian coordinate is 6.1707475808789 and body mass is 3519.72:

Xc = (6.1707475808789(1418.71 kg - ?kg/m^3))/?

Xc = ?

-No air space mass (zero-density volume).

**Torso segment part 4 (Bates et al., 2009 numbers):

Cartesian coordinate is 6.1433762338784 and body mass is 3415.46:

Xc = 6.1707475808789(1324.80 kg - 760.07 kg/m^3))/3547.84

Xc = (6.1707475808789(564.73))/3415.46

Xc = 3484.80628135/3415.46

Xc = 1.02030364324 m.

*Torso segment part 5 (Bates et al., 2009 numbers):

Cartesian coordinate is 6.9071243 and body mass is 3415.46:

Xc = 6.9071243(1324.80 kg - 760.07 kg/m^3))/3547.84

Xc = (6.9071243(564.73))/3415.46

Xc = 3,900.660305939/3415.46

Xc = 1.142060017080862 m.

**Skull-hip segment (Henderson, 2018 numbers):

Cartesian coordinate is 6.160045582537 and body mass is 3547.84:

Xc = (6.160045582537(2,190.45 kg - 850 kg/m^3))/3547.84

Xc = 8257.23310111/3547.84

Xc = 2.32739726175 m.

-Add up the weights from p. 6.

-Can’t add up Cartesian numbers, so don’t use Cartesian number from top row on p. 4.

Using the regular CoM/CoG equation, and the equation from Ibrahim et al., (2020b), I got a CoM ranging from 1.02-3.43 m. 1.02 m, 1.201 m, 2.33 m, and 3.43 m are the best estimates so far. These numbers are larger than the femoral length (62.5 cm; Ibrahim et al., 2020b, Supplementary Materials: Data File 2, Body dimensions, body mass, body segment masses, and whole body center of mass, p. 1), so Spinosaurus was definitely a quadruped. In fact, 1.02 m and 1.201 m fall around the 1.04-m estimate given by Ibrahim et al., (2014).

Update (9/6/23):
I tried to find a CoM equation in Sereno et al., (2020), but I didn't see one. Therefore, I'm sticking to the results that I got above for Spinosaurus' CoM.

Conclusions:

Based on my mathematical calculations, Spinosaurus' humerus was 80-90% the length of the femur. The neotype's was 81.1-92.6%, while Alpha Male 9109's was 92.8%. If the neotype's humerus was 51 cm, then the percentage would be 81.1-83.6%. All of these percentages are larger than a typical quadrupedal dinosaur like Stegosaurus (43-55%). The CoM for Spinosaurus ranges from 1.02-3.43 m for the animal (1.02 m, 1.201 m, 2.33 m, and 3.43 m, are the best estimates). This indicates that Spinosaurus was definitely a quadrupedal animal. Of course, it has been stated that theropods couldn't pronate their hands, preventing their palms from touching the ground. On the other hand, several trace fossils seem to indicate that theropods could've done some kind of pronation with their hands. Spinosaurus probably did minimal pronation, or even full pronation, with its hands. The possible manual ungual that might've belonged to Spinosaurus, NMC 41820, is less recurved than Baryonyx's and Suchomimus'. This suggests that Spinosaurus probably didn't use its hands in a similar fashion that the other two taxa did, and could have been for stability while walking.

I have come to believe that dinosaurs would've had any characteristic to help them survive in their habitats, regardless of what trait/characteristic we think they did, or didn't, have. I'm going to go on a limb here and say that Spinosaurus' hands were capable of pronation, and that the animal was a quadruped.

Links:
Brief History:
Stromer (1936) (p. 65):
https://www.zobodat.at/pdf/Abhandlungen-Akademie-Bayern_NF_33_0001-0102.pdf

Sereno et al., (1998) (p. 1300):

http://science.sciencemag.org/content/282/5392/1298
Link 2:
https://www.researchgate.net/publication/235242262_A_Long-Snouted_Predatory_Dinosaur_from_Africa_and_the_Evolution_of_Spinosaurids

Heckt (1996):

https://www.newscientist.com/article/mg15020312-900-science-how-to-give-a-dinosaur-the-hump/
Black (2011):

https://www.smithsonianmag.com/science-nature/was-spinosaurus-a-bison-backed-dinosaur-12849430/

Donald F. Glut (2001) (pp. 82 and 84):

https://books.google.com/books?id=O5znudeYRzEC&pg=PA82&lpg=PA82&dq=spinosaurus+in+books+1970&source=bl&ots=iscL5W5kaU&sig=ACfU3U2fJ9DJqndsOa7nXZbnpyRCZ88YOw&hl=en&sa=X&ved=2ahUKEwi16tiasonqAhU3RzABHVGFBsAQ6AEwFnoECAgQAQ#v=onepage&q=spinosaurus%20in%20books%201970&f=false

Ibrahim et al., (2014):
https://www.researchgate.net/publication/265553416_Semiaquatic_adaptations_in_a_giant_predatory_dinosaur
Supplementary Materials:
http://science.sciencemag.org/content/suppl/2014/09/10/science.1258750.DC1/Ibrahim.SM.pdf

Nash (2014) ("Did Bakker Get Spinosaurus Right After All?"):

https://antediluviansalad.blogspot.com/2014/08/did-bakker-get-spinosaurus-right-after.html

Henderson, Donald M. (2018):